Sep 12, 2014 · 5 minutes

A few years ago, Graham Dodge woke up with a terrible stomachache. Was it something he ate? A virus going around? He scoured the Internet for local disease data that might shed some light on his affliction, but had little luck. What data he could find was usually published by public offices only after an outbreak occurred.

Having exhausted his attempts and still suffering in pain, Dodge hopped on Facebook and saw that one of his friends complained about identical symptoms. And unless they had both eaten at the same restaurant, Dodge could deduce that it was a communicable illness that was going around town and therefore take whatever steps possible to protect his three children from contracting it.

And thus the idea for Sickweather was born.

Sickweather is an app that aggregates social media posts in your area to chart and forecast the spread of diseases. Dodge likens it to a weather app a person can check before leaving the house in the morning. Is it going to rain today? Better bring an umbrella. Are there a ton of people in the neighborhood with colds today? Better take extra care to wash your hands.

Dodge is obviously passionate about the idea and, as it turns out, so are investors. He tells me that Sickweather just closed a $350,000 round on AngelList though Foundry Group's FG Angel syndicate.

One of the keys to Sickweather's success is that it's built on data people are already sharing. But the biggest concern is, are the reports accurate? Recall earlier this year when an audit of Google's much-vaunted Flu Tracker revealed that it vastly overestimated the number of flu incidences.

Dodge insists that the problem with Google Flu Trends was it assumed that just because a person searched for flu symptoms, it must mean they have the disease. Sickweather's algorithm, on the other hand, only catches explicit statements like "I have the flu" or "My coworker has the flu."

"I think the main difference between what Google is doing and what we're doing involves 'explicit data' versus 'implicit data,'" Dodge says. "Google's data is very much implicit. It implies that these people have the flu because of looking up flu terms."

Even then, people may tweet that they have the flu when in reality they have the common cold (or common hypochondria). But Dodge says the impact of false positives is minimal.

"In aggregate, a lot of those false positives fall to the wayside," Dodge says.

As proof, Dodge claims Sickweather was able to predict the start of the 2012 flu season six weeks before the CDC confirmed it. It also identified whooping cough outbreaks several weeks before they were reported it the media, he says.

Some worry that when health and fitness apps struggle to gain traction it's because they require a behavioral change on the part of the user. In short, apps should be designed for the person we are, not the person we want to be. And while users already share personal disease data on social media, convincing them to use an app like Sickweather and take actionable steps to avoid contracting these illnesses may be easier said than done. After all, less than half of Americans over the age of 18 get vaccinated for the flu, despite easy and affordable access to shots.

But Dodge cites a few use cases where the app doesn't so much spawn a behavioral change as it does make existing behaviors easier:

"A mom told us that she received a strep throat Sickweather alert on her phone, and her elementary-aged son was complaining about a sore throat. This prompted her to take him into a swab test while she had the time to do that. He tested positive for strep and got him on regimen he needed to get back to school right away. She didn't have to miss any work."

In another example, an allergy alert prompted a parent to refill her children's prescription allergy medication so they would have it as soon as they developed symptoms.

"It's a little like the weather," Dodge says, emphasizing that the use of "weather" in the app's name is no coincidence. "People used to say, 'Why do I need a weather forecast when I can just look outside the window to see if it's raining or not?' There's a similar value proposition. People had never considered how valuable that information can be."

Seth Levine of Foundry Group, which runs the FG Angel syndicate through which Sickweather raised its seed round, agrees.

"They're basically creating the doppler radar of sickness," he says.

Levine also offered some advice to other startups looking to raise money through AngelList Syndicates.

"Simply posting your company on AngelList and waiting for the money to roll in isn't how it works. Companies need to do their own marketing and outreach (both on and off AngelList) -- that's not AngelList's job and the platform doesn't work (and isn't designed to work) that way. For companies that are willing to put in the effort it's a great way to coalesce investor interest."

As for how Sickweather will bring in revenue, it looks to target ads from health care providers and pharmaceutical companies based on the illnesses in a user's area. That may rub some the wrong way, but Dodge emphasizes that it will be very clear that these are advertisements, not medical advice.

Sickweather is still gaining traction, hitting 270,000 monthly unique visitors over the Summer. As part of Techstars' Sprint Mobile Heath Accelerator program, Sprint created a Sickweather widget and pre-loaded it on all its Android phones, helping to popularize the company. Unfortunately, Sickweather launched only on iOS, but its Android offering is set to launch in the next week or so, allowing the Sprint partnership to pay far greater dividends.

And then there's the fact Dodge predicts this year is going to bring another early and intense flu season. And while in a strange way that's a boon for Sickweather, it also means there's a huge opportunity for the app to make a difference in people's lives over the coming months.

[photo via US Army Corps of Engineers]